TikTok Coordinated Detection Project
1 Introduction
This document serves as a proof of concept, developed within the framework of the vera.ai Horizon EU project. It presents a comprehensive methodology for tracking and analyzing coordinated sharing activities on TikTok, employing traktok for data collection via the TikTok Research API and CooRTweet for the analysis of coordinated behavior patterns.
Initial Discovery: The analysis initiates by focusing on content tagged with the #moskow hashtag. This first step successfully identifies a preliminary group of accounts involved in coordinated sharing activities in the aftermath of the attack in Moscow. This discovery lays the groundwork for an extensive examination of coordinated dynamics across social media platforms.
On April 19, 2024, we received a list of 513 problematic accounts from a trusted partner within the vera ai consortium. Accounts that were mentioned at least twice on this list have been added to our pool of monitored accounts.
Daily Monitoring and Analysis: Subsequent to the initial identification, the methodology transitions into a phase of daily monitoring. In this phase, the script consistently retrieves videos posted by the previously identified accounts, with the goal of detecting both ongoing and emerging instances of coordinated behavior. As new accounts manifesting coordinated behavior (time_window = 120 seconds, min_participation = 2 posts, edge_weight = 0.9) are discovered, they are incorporated into the daily monitoring routine. For each identified network of coordinated accounts, AI-generated labels are automatically created using the gpt-4o model by analyzing the content patterns, themes, and characteristics of the posts shared within each network, providing a concise description of the common elements that unite these accounts in their coordinated behavior.
This approach ensures continuous updates on the number of newly discovered coordinated accounts, highlighting the fluid nature of social media coordination. Enhanced by interactive visualizations, the analysis sheds light on the shifting landscape of account activities and the intricate network of interactions among them on the TikTok platform.
By delineating these processes, the proof of concept underscores the potential for advanced analytical tools to reveal and understand the complex phenomena of coordinated social media behavior within the context of significant societal events.
2 Today Results
Execution Summary: We attempted to retrieve videos from 6,844 monitored accounts, during the period from July 29, 2025 to August 05, 2025. We successfully retrieved 15,182 recent videos. Using available data, we accessed a total of 15,161 videos posted on TikTok within the timeframe. Our analysis for coordinated detection in these videos identified 926 accounts spread across 207 components, and it also uncovered 83 new accounts exhibiting coordinated behavior.
::: {.callout-warning}
Warning: 206 errors were encountered during execution. Results may be incomplete. Please check the error summary below for details.
:::
Error Details
Total errors encountered: 206
2.0.1 Errors by type:
chunk_error daily_quota_exceeded empty_chunk
175 1 30
2.0.2 Sample Errors:
- Type: chunk_error, Message: HTTP 500 Internal Server Error. • status: internal_error • message: Something is wrong. Please try again later. • log_id: 202508050228120793375136F6D30AE2D6
- Type: chunk_error, Message: HTTP 500 Internal Server Error. • status: internal_error • message: Something is wrong. Please try again later. • log_id: 202508050127255B7F6FCA2E508D059292
- Type: chunk_error, Message: HTTP 500 Internal Server Error. • status: internal_error • message: Something is wrong. Please try again later. • log_id: 20250805040434F8DF86075692AB011C1F
- Type: chunk_error, Message: HTTP 500 Internal Server Error. • status: internal_error • message: Something is wrong. Please try again later. • log_id: 20250805025319AB6053FC1C4236011A45 character(0)

2.1 Interactive Graph Visualization
2.2 Detailed Tables
2.2.1 Coordinated Accounts
2.2.2 Coordinated Networks/Clusters
2.2.3 Coordinated Posts
Warning in instance$preRenderHook(instance): It seems your data is too big for
client-side DataTables. You may consider server-side processing:
https://rstudio.github.io/DT/server.html
Download Results
Download a comprehensive package of today's coordination analysis results, including:
- Coordinated accounts table (CSV format)
- Network clusters analysis (CSV format)
- Coordinated posts data (CSV format)
- Network graph (GraphML format)
- README file with detailed field descriptions
3 About
| vera.ai is a research and development project focusing on disinformation analysis and AI supported verification tools and services. Project funded by EU Horizon Europe, the UK’s innovation agency, and the Swiss State Secretariat for Education, Research and Innovation |
4 References
Giglietto, F., Marino, G., Mincigrucci, R., & Stanziano, A. (2023). A Workflow to Detect, Monitor, and Update Lists of Coordinated Social Media Accounts Across Time: The Case of the 2022 Italian Election. Social Media + Society, 9(3). https://doi.org/10.1177/20563051231196866
5 Session Info
R version 4.3.3 (2024-02-29)
Platform: x86_64-pc-linux-gnu (64-bit)
Running under: Ubuntu 22.04.4 LTS
Matrix products: default
BLAS: /usr/lib/x86_64-linux-gnu/blas/libblas.so.3.10.0
LAPACK: /usr/lib/x86_64-linux-gnu/lapack/liblapack.so.3.10.0
locale:
[1] LC_CTYPE=C.UTF-8 LC_NUMERIC=C LC_TIME=C.UTF-8
[4] LC_COLLATE=C.UTF-8 LC_MONETARY=C.UTF-8 LC_MESSAGES=C.UTF-8
[7] LC_PAPER=C.UTF-8 LC_NAME=C LC_ADDRESS=C
[10] LC_TELEPHONE=C LC_MEASUREMENT=C.UTF-8 LC_IDENTIFICATION=C
time zone: Etc/UTC
tzcode source: system (glibc)
attached base packages:
[1] stats graphics grDevices utils datasets methods base
other attached packages:
[1] htmltools_0.5.7 DT_0.33 traktok_0.0.7.9000 logger_0.3.0
[5] progress_1.2.3 jsonlite_2.0.0 visNetwork_2.1.2 igraph_2.1.2
[9] scales_1.3.0 httr_1.4.7 CooRTweet_2.1.0 lubridate_1.9.4
[13] forcats_1.0.0 stringr_1.5.1 dplyr_1.1.4 purrr_1.0.4
[17] readr_2.1.5 tidyr_1.3.1 tibble_3.2.1 ggplot2_3.5.1
[21] tidyverse_2.0.0
loaded via a namespace (and not attached):
[1] gtable_0.3.4 bslib_0.6.1 xfun_0.42
[4] httr2_1.1.0 htmlwidgets_1.6.4 tzdb_0.4.0
[7] crosstalk_1.2.1 vctrs_0.6.5 tools_4.3.3
[10] generics_0.1.3 curl_6.2.1 parallel_4.3.3
[13] pkgconfig_2.0.3 data.table_1.16.4 tidytable_0.11.2
[16] lifecycle_1.0.4 compiler_4.3.3 farver_2.1.1
[19] munsell_0.5.0 sass_0.4.8 yaml_2.3.8
[22] jquerylib_0.1.4 pillar_1.10.1 crayon_1.5.3
[25] ellipsis_0.3.2 openssl_2.3.2 cachem_1.1.0
[28] tidyselect_1.2.1 digest_0.6.34 stringi_1.8.4
[31] labeling_0.4.3 fastmap_1.2.0 grid_4.3.3
[34] colorspace_2.1-0 cli_3.6.4 magrittr_2.0.3
[37] RcppSimdJson_0.1.11 withr_3.0.2 prettyunits_1.2.0
[40] rappdirs_0.3.3 bit64_4.0.5 timechange_0.3.0
[43] rmarkdown_2.25 bit_4.0.5 askpass_1.2.1
[46] hms_1.1.3 evaluate_0.23 knitr_1.45
[49] rlang_1.1.5 Rcpp_1.0.14 glue_1.8.0
[52] vroom_1.6.5 R6_2.6.1